Neural geometry and excitability Romain Brette romain.brette@inserm.fr

Classic theory of excitability Na+ Na+ K+ K+ The squid giant axon A model of excitation of a homogeneous membrane Hodgkin & Huxley (Nobel prize 1963) axon cell bodies

Geometry of spike initiation in neurons soma AIS Kole & Stuart (2008) ~ 40 µm What is different about spike initiation in neurons is not the ionic basis but the geometry of the system. Synaptic currents coming from the dendrites are collected at the soma, which increases the potential of the soma and neighboring regions, including a small structure in the axon near the soma called the axonal initial segment, a few tens of µm long. That region is packed with ionic channels, in particular Na+ channels, and this is where the action potential first occurs, before it is seen in the soma, where the spike is actively regenerated by local channels and spreads to the dendrites. Electrically this region is very small because the characteristic distance of passive voltage attenuation is several hundred µm. So this is an electrical system characterized by a discontinuity in the geometry and distribution of channels, quite unlike the squid giant axon. Recently it has been found that the both the position and length of this initial segment varies with activity. It also varies between cells, with development, and with diseases. AIS geometry is plastic Hamada et al. (2016) Grubb & Burrone (2010)

Geometry of spike initiation in neurons What is the impact of geometry on excitability? on spike initiation on spike transmission to soma What is the impact of geometry on excitability ? How does it impact the initiation of spikes, in particular the threshold ? How does it impact the transmission of the spike to the big soma ? How does it impact the flow of ions, which determines energy consumption ? These are the main questions we want to answer in this project.

Resistive coupling theory in a nutshell AIS soma 1. Spike initiation resistive capacitive resistive = Na+ Na+ ~ d2/x ~ exp(V/k) resistive d x 2. Spike transmission resistive = capacitive ~ d2/x ~ soma area . dV/dt The initial approach I have taken is to model the soma and AIS as a dipole. At spike initiation, a current loop forms where current enters the axon carried by Na+, flows resistively towards the soma, and escapes as capacitive current. From this simple model, we can get a sense of the impact of geometry on electrical properties, in particular on the threshold for spike initiation. The resistive current should be proportional to the section area of the axon and inversely proportional to the distance of the initiation site from the soma. The Na+ current is an exponential function of voltage. From this, we deduce a logarithmic dependence between threshold and these geometrical parameters. In the same way, for the transmission to the soma, the capacitive current is proportional to the membrane area of the soma and dendrites, and for an action potential of given shape, the distance of the initial segment should be inversely proportional to the area of the soma. This is a good start but of course the initial segment is not a point. So we will need to analyze spike initiation and spike transmission in spatially extended models based on cable theory, using separation of time scales to obtain analytical formulae. We will also analyze the effect of channels at the initial segment on excitability. Finally, in contrast with homogeneous axons, there are significant ion flows in the axial direction, for example Na+ flowing to the soma at initiation. Cable theory only includes axial currents due to the electrical field but not to diffusion and so we will look at an extended version of cable theory. From this analysis, we will derive the total influx of ions due to action potentials, which determines energy consumption. And finally, this will allow us to deduce the optimal properties of the initial segment for minimum energy consumption.

Resistive coupling theory in a nutshell Simulation of a detailed model (Yu et al. 2008) Teleńczuk M, Fontaine B, Brette R (2016). The basis of sharp spike onset in standard biophysical models.

Spike initiation in the AIS Most of the time, Vm is the same in soma and AIS Stuart, Schiller, Sakmann (J Physiol 1997) Kole, Letzkus, Stuart (Neuron 2007) INa x ≈ 30 µm conductance ratio: λ/x ≈ 15 soma = « current sink » soma Ra = x.ra Ra = λ.ra Brette, R. (2013). Sharpness of spike initiation in neurons explained by compartmentalization.

Spike initiation in the AIS INa x ≈ 30 µm conductance ratio: λ/x ≈ 15 soma = « current sink » soma Ra = x.ra Ra = λ.ra Brette, R. (2013). Sharpness of spike initiation in neurons explained by compartmentalization.

Effect of channel location The soma is a current sink soma Va Ra.I (Ohm’s law) Vs

Effect of channel location The soma is a current sink soma Va Ra.I (Ohm’s law) Vs

Resistive current = Na+ current Na activation I=f(Va) soma Va Ra.I Vs

Resistive current = Na+ current I=f(Va) I=(Va-Vs)/Ra Lateral and Na currents must match soma Vs Vs I=f(Va) I (nA) COLOR CODES!!!! I=(Va-Vs)/Ra Bifurcation! = abrupt opening of Na channels

A view from the soma Lateral current flows abruptly when a voltage threshold is exceeded m Na channels open in an all-or-none fashion « sharp initiation »

A view from the soma A fairly good phenomenological description: below Vt, no sodium current when Vm reaches Vt: all channels open (spike) a.k.a. the integrate-and-fire model ! with axonal initiation single compartment HH model Vt m Brette R (2015). What Is the Most Realistic Single-Compartment Model of Spike Initiation?

The threshold equation Fixed point equation I (nA) f(Va) = (Va-Vs)/Ra Spike threshold = bifurcation point (= Vs when solution jumps) Na activation The threshold equation A small correction actually ka V1/2

Experimental evidence 1) Dipole formation depolarization Ve AIS Vm soma Palmer and Stuart (J Neurosci 2006) Chorev & Brecht (J Neurophy 2012) repolarization Ve soma Hippocampal neurons Teleńczuk M, Fontaine B, Brette R (2016). The basis of sharp spike onset in standard biophysical models.

Experimental evidence 2) Sharp spike initiation Somatic voltage-clamp recordings of Na current Sharp spike onsets Fast response to current changes, as in integrate-and-fire models! Axonal Na channels inactivated Milescu, Bean, Smith (2010) Sharp I-V relationship at soma Ilin et al. (2013) V Naundorf, Wolf, Volgushev (2006) Badel et al. (2008)

Experimental evidence 3) Threshold The threshold equation Axonal threshold Va = Vs + ka (Fekete and Debanne, preliminary data) Kole & Stuart (2008)

Experimental evidence 4) Threshold The threshold equation Pinching (Fekete and Debanne, preliminary data)

Spike transmission to the soma Yu et al. (2008) The AIS charges the soma

Resistive coupling theory, part 2 Theory: resistive current (axon) = capacitive current capacitive resistive x Resistive current = 1/R ~ 1/x Capacitive current ~ somatodendritic area . dV/dt Conclusion (for a spike of fixed shape): x ~ 1/somatodendritic area A big neuron must have a proximal AIS

Neuron size vs. AIS position Big neuron => proximal AIS Initial AP shape is normalized Hamada, Goethals, de Vries, Brette, Kole (PNAS 2016). Covariation of axon initial segment location and dendritic tree normalizes the somatic action potential.

Resistive coupling theory We neglect inactivation etc clamped at threshold resistive all channels open Rough approximation: I = (ENa – Vthreshold) / (ra.Δ) More precisely Result: with

Resistive coupling theory Voltage response of a cylindrical cable (Rall): Normalized voltage response =>

Theory vs. experiment

Summary: resistive coupling theory AIS soma 1. Spike initiation resistive capacitive resistive = Na+ Na+ ~ d2/x ~ exp(V/k) resistive d x Sharp initiation Threshold equation 2. Spike transmission resistive = capacitive co-tuning of AIS distance and neuron size ~ d2/x ~ soma area . dV/dt

Thank you Maria Teleńczuk Jonathan Platkiewicz Maarten Kole Sarah Goethals Dominique Debanne Catherine Villard Bertrand Fontaine Hamada M, Goethals S, de Vries S, Brette R, Kole M (2016). Covariation of axon initial segment location and dendritic tree normalizes the somatic action potential. Teleńczuk M, Fontaine B, Brette R (2016). The basis of sharp spike onset in standard biophysical models. Brette R (2015). What Is the Most Realistic Single-Compartment Model of Spike Initiation? Brette R (2013). Sharpness of spike initiation in neurons explained by compartmentalization. Platkiewicz J, Brette R (2011). Impact of Fast Sodium Channel Inactivation on Spike Threshold Dynamics and Synaptic Integration. Platkiewicz J, Brette R (2010) A Threshold Equation for Action Potential Initiation.